Operations for Conceptual Schema Manipulation: Definitions and Semantics

Abstract

This paper concerns the issue of conceptual schema integration and manipulation, i.e., the process of manipulating highly abstract, semantically rich database schemas for the extraction of meaningful and unambiguous results. We propose the use of six fundamental manipulation operations, namely rename, select, project, union, set difference, and intersection. We give their definition and semantics using the Entity-Relationship modeling notation. We argue that in order to preserve the semantics of the ER-schemas before and after the operations are performed, these schemas need to be translated to mathematical formulations; then the manipulation operations can be applied. We use the $\mathcal{A}\mathcal{L}\mathcal{C}\mathcal{Q}\mathcal{I}$ description logic notation which features a rich combination of constructors, powerful enough to express elements of ER-schemas. We show that the resulting knowledge bases, which encapsulate all the knowledge about the ER-schemas, need further structuring to accommodate the semantics of the ER elements. Examples demonstrate the applicability and efficiency of the proposed approach.